Ask Your Question
0

diagnostic message queue among nova-api and nova-compute

asked 2015-01-07 07:21:02 -0500

TristanSu gravatar image

I had icehouse installed on my CentOS and now I'm debugging an issue in nova. The nova service are all in down state in the output of "nova service list" and it looks that the nova-api could not talk with nova-compute. Is there any way to diagnostic the message exchanged between nova-api and nova-compute? I am not able to find any tool to monitor this. I use the default rabbit mq.

any help is appreciated.

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted
0

answered 2015-01-07 23:54:22 -0500

TristanSu gravatar image

updated 2015-01-07 23:57:30 -0500

after a night of investigation, I figured out the root cause. It was caused by rabbitmq which blocked all message publishing due to the database space limit.

=WARNING REPORT==== 5-Jan-2015::09:20:03 ===
disk resource limit alarm set on node 'rabbit@ngncloud-controller'.

**********************************************************
*** Publishers will be blocked until this alarm clears ***
**********************************************************

rabbitmq provides a management plugin which can be used to view the queue status and other statistics. It's easy to set up but requires me to add iptables rule for allowing port 15672 manually, to make the web interface work.

hope this would be helpful for others.

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2015-01-07 07:21:02 -0500

Seen: 127 times

Last updated: Jan 07 '15